1. Field of the Invention
This invention relates to a surround headphone output signal generator for use in multichannel reproducing apparatus to combine multichannel signals into two-channel signals and output the two-channel signals to headphones.
The present application claims priority from Japanese Application No. 2002-135862, the disclosure of which is incorporated herein by reference.
2. Description of the Related Art
Multichannel reproducing apparatus such as that in an A/V amplifier, a DVD player and the like typically generates source (sound) signals for right and left front channels, a center channel, right and front surround channels, and additionally a subwoofer channel, from digital audio data stored on a DVD (Digital Versatile Disc), a CD (Compact Disc) or the like recorded through multiple channels, and outputs the generated signals to five or more speakers placed around the listener, in order to form a sound field in three-dimensional sound which gives the listeners the feeling of being at a live performance or a theater.
Listening to the three-dimensional sound, reproduced by the multichannel reproducing apparatus, through headphones requires combining the source signals for the respective channels generated from the digital audio data to re-generate source signals for the two channels.
Such combining of the multichannel source signals into the two-channel source signals is referred to as stereo-downmixing.
FIG. 2 is a block diagram illustrating the conventional configuration for executing stereo-downmixing in the multichannel reproducing apparatus incorporating headphone terminals.
In FIG. 2, a digital-audio interface receiver (hereinafter referred to as “DIR”) 1 receives the input of two-channel digital audio data which is transmitted from digital audio equipment of various types into the multichannel reproducing apparatus.
The DIR 1 carries out demodulation to extract a clock and data from the incoming digital audio data, and then outputs the resulting clock (CK) and data (DATA) to a multichannel decoder 2.
Then, the multichannel decoder 2 generates channel signals (5.1 channel in this case) consisting of a left front signal L, a right front signal R, a center signal C, a left surround signal LS, a right surround signal RS and a sub-woofer signal LFE, from the data (DATA) supplied from the DIR 1 in accordance with the associated clock (CK), and then outputs the resulting signals to the stereo-downmixing processor 3.
The stereo-downmixing processor 3 combines the left front signal L, the right front signal R, the center signal C, the left surround signal LS, the right surround signal RS and the sub-woofer signal LFE on the basis of preset downmixing coefficients, to produce two-channel surround headphone output signals Lo and Ro for outputting to the headphones.
The following is the mathematical expressions of representing an example of the combining process for the stereo-downmixing signals in the stereo-downmixing processor 3:Lo=L+0.707C+LS+LFE Ro=R+0.707C+RS+LFE 
As seen from the mathematical expression, in the process of combining into the surround headphone output signals Lo and Ro, if the source signals for the respective channels to be combined are concurrently increased in level, the gain to the surround headphone output signals Lo and Ro may exceed one (zero dB), which causes clipping in the sound reproduced from the headphones.
For this reason, in order to prevent clipping in the sound reproduced from the headphones, the prior art performs a multiplication by an attenuation coefficient α in the process of combining into the surround headphone output signals Lo and Ro as follows:Lo=α (L+0.707C+LS+LFE) Ro=α (R+0.707C+RS+LFE). 
The attenuation coefficient α is determined taking into consideration a degree of margin and risk because a small value given to the attenuation coefficient α (i.e. a high attenuation rate) reduces each level of the surround headphone output signals Lo and Ro so as to result in a decline in an S/N ratio, whereas a large value (i.e. a low attenuation rate) makes it impossible to prevent clipping in the sound reproduced by the surround headphone output signals Lo and Ro.
However, the balance between the recording levels of the respective channels in the multichannel source varies among the types of multichannel sources.
For example, in the case of a movie source such as a DVD and the like, the LFE channel or the surround channel is increased in level, or alternatively all channels are increased in level in an instant for the purpose of producing dramatic sound effects. In the case of a music source such as a CD, an audio DVD and the like, the front channels (L, C and R) have a high level because the sound of the instruments played or the vocal performed on the stage is concentrated at the front, but the surround channel for reproducing reflected sound or sound generated from the audience for the purpose of creating the atmosphere in a concert hall or the like has a relative low level, and the LFE channel is less used.
For these reasons, conventionally, there are some problems: even if an attenuation a is set at any of the values in the stereo-downmixing process, it is impossible to completely prevent clipping in the sound reproduced from the surround headphone output signals Lo and Ro in the case of a movie source, and to prevent a decline in the S/N ratio in the case of a music source.